Study Shows: ChatGPT Achieves Almost 50% Score on Board Certification Practice Test for Ophthalmology

Date:

A new study published in JAMA Ophthalmology has found that the artificial intelligence system known as ChatGPT scored just 46% on a board certification practice test for ophthalmology. The performance improved to 58% the second time it was tested, one month later.

The examination used a dataset of practice multiple choice questions from a commonly used resource for board certification exam preparation, and the response was at a great convenience since ChatGPT is free and accessible from any internet connection.

The use of artificial intelligence (AI) in medical and exam practice has become increasingly popular since its launch in November, 2022, but many are concerned that the accuracy may not be reliable.

The study was led by Rajeev H. Muni from St. Michael’s Hospital, a site of Unity Health Toronto. He has stressed for the responsible use of such AI systems, and recommended that ChatGPT is not suitable for providing assistance in preparing for board certification at the moment.

The research team analysed 125 text-based multiple-choice questions and provided feedback on the results. When comparing the outcomes with the most popular responses of ophthalmology trainees, the AI system answered 44% of the questions correctly and was least popular 11% of the time in selecting the appropriate answer.

Another researcher, Marko Popovic of the Department of Ophthalmology and Vision Sciences at the University of Toronto, is optimistic on the role of AI in medical education. He believes that with the development of AI, it will eventually be able to provide more exact answers.

The most accurate answers from ChatGPT were regarding general medicine questions which quizzed 79% correctly. The chatbot scored the lowest in niche subspecialties such as oculoplastics (20%) and retina (0%).

See also  Hack ChatGPT and Go Grandma Mode with These Easy Tips

Andrew Mihalache, a Western University undergraduate student and lead author of the study, foresees the potential for ChatGPT accuracy to improve over time. As such, it is important to ensure responsible use and ensure that it does not provide any short cut for board certification.

Frequently Asked Questions (FAQs) Related to the Above News

Please note that the FAQs provided on this page are based on the news article published. While we strive to provide accurate and up-to-date information, it is always recommended to consult relevant authorities or professionals before making any decisions or taking action based on the FAQs or the news article.

Share post:

Subscribe

Popular

More like this
Related

AI Films Shine at South Korea’s Fantastic Film Fest

Discover how AI films are making their mark at South Korea's Fantastic Film Fest, showcasing groundbreaking creativity and storytelling.

Revolutionizing LHC Experiments: AI Detects New Particles

Discover how AI is revolutionizing LHC experiments by detecting new particles, enhancing particle detection efficiency and uncovering hidden physics.

Chinese Tech Executives Unveil Game-Changing AI Strategies at Luohan Academy Event

Chinese tech executives unveil game-changing AI strategies at Luohan Academy event, highlighting LLM's role in reshaping industries.

OpenAI Faces Security Concerns with Mac ChatGPT App & Internal Data Breach

OpenAI faces security concerns with Mac ChatGPT app and internal data breach, highlighting the need for robust cybersecurity measures.